A Semi-Empirical and a Neural Network Model of a Batch Plant Dynamics for Flexible Recipes Control

Abstract In order to apply a model based control strategy such as flexible recipes control to an industrial batch process, a model of the process dynamics is needed. This paper presents the modelling procedure for two such models: a semi-empirical and an artificial neural network model. Both models predict precipitation rates of TiO 2 particles in an industrial hydrolysis process. Model properties and their prediction accuracy is compared. The artificial neural network is trained using the augmented training data set approach. A simulator has been designed to study the application Of flexible recipe instructions.